AIM RE-bal

[nitelord]

Hi Matt -

Doesn't it seem unusual that your inverse stock underperformed the lichello series stock when AIMed? Logic seems to suggest that performance of both should have been comparable.

John

[2mc]

Hi Jibes,

>Now let me get this straight. You did just a re-balance on the two without using AIM as in AIM Re-Bal to get the $2.68mil?

Yep, that's right.

I'll be doing some other tests as I think through the process. I mentioned some tests I going to do in my last post. If anyone wants to help me think through this and give me ideas (other than the ones I have already stated) on how to set up some tests that will prove or disprove the idea, then let me know. Keep in mind, I won't be open to optimization kinds of tests (just my quirk - hope it doesn't offend anyone.)

I also may not be as fast as everyone would like, because I am working on several projects at the same time. But, it is in my own self-interest to pursue this so I will.

Matt

[jibes]

karel,

No problem here, when AIM Re-Bal doesn't prove to be the silver bullet of investing! As long as we have a fair idea what kind of stock and portfolio composition is necessary, things are OK

Thanks for your encouragement.

The more I think about how it works the more sense it makes. I was a little set back by the results because I feared that I was jumping the gun on AIM RB since I didn't have much data to support my claim that this thing worked. But because of some of the comments and testing on Matt's modifications I can see that stock selection can (at least in this case) have an influnce on a system.
I will have to do more testing though to see how well this AIM Re-Bal really works.

Jibes
AIM Re Bal at:
http://jibes0.tripod.com/trendseeker.html

[jibes]

Matt.

This is great! I wanted to do a study using the Lichello series but I didn't quite know how to go about it. But inverting one against the other would work fine.

Now let me get this straight. You did just a re-balance on the two without using AIM as in AIM Re-Bal to get the $2.68mil?
If that is true then maybe Re-Bal would be quite powerfull on the right kind of stocks. In the past I tried it on the Dow 30 stocks and wasn't impressed. Why I didn't try it on some wilder stocks beats me.

Great work Matt!

Jibes
AIM Re Bal at:
http://jibes0.tripod.com/trendseeker.html

[karw]

Hi Matt, weighted average is the average price on the LIFO stack.

WAP= (P1*Q1+P2*Q2+.....Pn*Qn)/(Q1+Q2+....Qn)

where P1 is the price of the first stack element, Q1 is the quantity of stock of the first stack element.

The formulas I corrected to equalize the impact of 4/9 vs 9/4 so that a percentage downwards move and the corresponding percentage upward move would result in the same amount. In this way you see that the system never asks for a sell greater than the quantity in stock. On the buy side it means you NEVER will be out of your cash.

The situation is fully symmetric and this is caused by the sell factor, which is a quadratic of a price ratio.

It will be interesting to see how a formula like this will perform in real life. No worries about deep divers!

Kind Regards, K

[2mc]

Hi K, good to see you applying yourself. Good work. Question: What's the weighted average? I gather you aren't using LIFO. And, your 'actual cost' - is it LIFO, FIFO, or average?

Regarding your formula, I'll need to look at it. I'll need to derive it mathematically. I'll give it a looksee. But, my first reaction is that what is done to the buy should be done to the sell. You know, "what's good for the goose is good for the gander."

Keep up the good work.

Matt

[karw]

Hi Matt,

Made a spreadsheet similar to your machine. Made a one stock stack(with multiple stocks would have had a tracker share as you described).
Used the portfolio Value and portfolio actual cost in the calculations.

Sell Amount = Value * Sell factor

Buy Amount = Cash * Buy Factor

Buy Factor = (1 - value/actual cost)

Weighted average price per share = WAP

Actual share price = AP

Sell factor = {(value/actual cost - 1) * WAP *WAP } / {AP*AP}

In this way the Buy and Sells are scaled to a similar amount for 4/9 and 9/4.

Please react, K

[2mc]

Hi Jibes,

Since you're rethinking the Rebal scheme, I thought I'd do the same. I did not fully perform my normal thought processes and checks before I decided to implement rebalancing. So, before my account suffers too much (if indeed it will suffer), I will do some investigation. I will share what I come up with. In these tests I will be using the standard AIM with my modification that I described on the AIM board. I will not be using what I had earlier discussed here. That will be for another day.

To that end I did an initial test. I wanted to test the max that rebalancing can do on its own. So, I set up 2 fake stocks - one with the Lichello series as its price history and one having the inverse.

I started with $10,000 to invest - $5000 in each stock. I rebalanced each month with no minimum investment. The result after 8 years is $2.68mil.

I then AIMed (with my modification) the two stocks separately - each having $2500 invested and $2500 in Cash. This totals $10,000 initial investment when you add the two together. The stock starting at $10 made $967,457 and the stock starting at $4 made $385,287 for a total of $1,352,744 - which is a little more than half what the fully invested 'rebalancing' portfolio made. (This seems to make sense, since it is generally 50% invested.)

Now, what remains to be seen is whether there can be a synergy between AIM and rebalancing or whether they mute each other's performance. We also need to see how rebalancing does on stocks that aren't *exactly* correlated.

I'll be posting the results of my tests as I think through the process.

Matt

[Qarel]

Hi Jibes,

No problem here, when AIM Re-Bal doesn't prove to be the silver bullet of investing! As long as we have a fair idea what kind of stock and portfolio composition is necessary, things are OK.

Regards,

Karel

[lostcowboy]

It may work better on the dogs of the Dow investment method.

[jibes]

To All...

Please go to my site and read the small mesage I posted there where it says "New note about AIM Re-Bal."

Thanks

Jibes
AIM Re Bal at:
http://jibes0.tripod.com/trendseeker.html

[jibes]

Am I correct in thinking that your re-bal calculation and your aim calculation go hand in hand. In other words you do both together. I thought I read a post where someone was going to do the re-bal calculation weekly and aim monthly(??).

Yes. the Re-Bal is done at the same time as you do AIM. The two do go hand in hand. Some months AIM says no buy or no sell. You will still do the Re-Balance and even that might say take no action. More than likely though it will take some action. I recomend doing it once a month.

However there are times when a stock in the port will fall a great deal (or rise). Lets say 15% in one or a couple of days during the middle of the month. I will then do a "special" It's just too good to pass up. What I do is do the AIM Re-Bal at that time instead of waiting till the end of the month. Of course this has it's dangers because how low is low and how high is high? The stock may still go down or up farther. Common sense should prevail.

Also, Let's say you had a portfolio of 4 stocks. When you do your re-bal calculation(along with AIM), you should get an exact buy or sell amt for each stock. So you don't have to decide which stock in your portfolio to apply your buy or sell to (which isn't the case when using AIM alone), re-bal tells you how much of each to buy or sell. Is that correct?

The formula shown on my website shows that each stock is compared to the "IDEAL VALUE." Any of the stocks will fall above or below this number and the diference is the amount you buy or sell. I ignore any order under $100 for the time being. I think is should be a min percentage though rather than a fixed amount.

Jibes

[2mc]

Ok, is there a "jibe" and a "jibes?" That's too confusing for me.

I rebalance weekly and "AIM" monthly. And, if I understood your question correctly, then I think you correctly stated it.

But, in reality for me (because I am with folioFN) I don't have to set exact prices, I can just tell folioFN: (when rebalancing only) "Reset my portfolio to the original percentages" and (when rebalancing and 'AIMing') "Here's what the new Stock Balance should be and reset to original percentages."

It's really easy at folioFN. Since my account is an IRA and I don't have any commissions (to speak of) and I can buy fractional shares, I don't worry about any minimums.

Matt

[jibe]

Questions

Am I correct in thinking that your re-bal calculation and your aim calculation go hand in hand. In other words you do both together. I thought I read a post where someone was going to do the re-bal calculation weekly and aim monthly(??).

Also, Let's say you had a portfolio of 4 stocks. When you do your re-bal calculation(along with AIM), you should get an exact buy or sell amt for each stock. So you don't have to decide which stock in your portfolio to apply your buy or sell to (which isn't the case when using AIM alone), re-bal tells you how much of each to buy or sell. Is that correct?

Thanks

[2mc]

I'm sorry I'm being as clear as mud. And, I have a degree in Communication! <lol>

You've almost got it. However, I do *not* keep the average true cost (LIFO) of each ETF. I only keep it on the group of ETFs (and since I have 2 groups, I have two average true costs to calculate). But, you may ask, what is a share of the group? Good question. Treat it like a mutual fund. You can arbitrarily set the shares of your mutual fund to anything you want - 100 shares, 1000 shares, 100000 shares. Your initial cost per share will be the total invested divided by this arbitrary amount.

Now, whenever you make a "purchase" - that is, you move money from cash to the invested side - the cost per share of this new transaction will be the current value of the investments divided by the number shares invested (that is, the number just prior to the new additions).

So, you ask how do I know what to buy of which ETF? Well, FolioFN helps me with that. I can set up weights for each ETF in my group. A weight is that percentage of the whole group that the ETF holds. Since I am always rebalancing, those percentages never change. So, I tell FolioFN: the total amount in the invested side should be $X and rebalance to original weights. If you're not in FolioFN you will have to make the calculations yourself. I don't have tax or commission concerns so I don't have to set a minimum. You may have to set a minimum trade.

Hope that makes things a little clearer.

Matt

[2mc]

Let me try to be a little clearer. I'm going to use the beginning part of your illustration to help make the point clearer - at least, I will attempt to make it clearer.

Your scenario:

Price...SharesBot....Cost
25......4............100
10......10...........100
4.......20...........100

A total of 34 shares were purchased for a total cost of $300. This means the true average cost is, as you said, $8.82 per share.

Now, forget for a moment how one determines the number of shares to sell, let's just say we want to sell 25 shares for whatever reason.

Starting from the last, we have 20 of the 25 shares available for sale, so we sell them. Then, we have 10 shares available for sale, of which we only need 5, so we sell 5 of those. This is how the table would look after this sale:

Price...SharesBot....Cost
25......4............100
10......5............50
4.......0............0

We now have 9 shares left at a cost of $150. Our true average cost *LIFO* is now $16.67 per share. So, after the sale the true average cost went up. I hope that is clearer.

Regarding whether or not Synchrovest uses average buying price or average true cost, I see your point. However, since Synchrovest sells 100% when there is a sale, the average buying price and the average true cost are the same. I guess I was referring to what one would do, if one were trying to operate an AIM program like Synchrovest, where there would be partial sales. Thanks for pointing out that I wasn't spot on accurate in that description.

Matt

[karw]

Hi 2mc,

Thanks for your program description.

I understand the average cost calculation(I think). You put the prices on a LIFO stack and when you sell you take some shares from the stack. The stock on the stack will be used for an average cost calculation.

Each sector fund has its own stack and its own calculation. True total cost is total of the stacks.

Understand that you wrote a program for this, this is not easily implementable in a spreadsheet. (I use a very, very simple limited stack for LIFO calculations in my AIM spreadsheets, and that was difficult to develop)

When you talked about the asymmetry in the Value/Cost ratios
you described a solution to this unwanted phenomenon. I assume that you based this solution on your Rebalancing experience. For me it is not transparent, so please can you elaborate a little bit more here.

I totally agree with your remarks about tweaking. Also the use of ETFs/sector funds is a very rational choice in my view. I hope my rebalancing experience will be as fruitful as yours.

Kind Regards, K

[karw]

Hi All,

Over the weekend I rebuilt my Portfolio spreadsheet to an AIM ReBAL spreadsheet.

Have divided 22 equities into 5 "sectors", each "sector" is 20% of total. Then play it like Jibes described. I will not use target ratios based on avg buying price for each stock, but take the sector approach.

Allocating stocks to sectors and rebalancing them over the weekend also provided the opportunity to sell 2 regio funds completely(Poland and Pacific Basin). I want less stuff in my portfolio and want an allocation similar to the misc sector program of 2mc. I will slowly rebuild my portfolio towards a well 'balanced' sector program.

Planning to rebalance weekly and AIM monthly.

The Synchrovest Biotechnology machine was taken out and will from now on exist independently from my ReBal program.

Best Regards, K

[lostcowboy]

Hi 2mc,
You said The crucial calculation is the true average cost which is based on LIFO. Do you mean something like this? (Don Carson asked me) This is part of my reply to him.

(Maybe, got to learn more about it. This may be what he is doing

Stock Monthly      Shares   Total Total  Average  
      Price installment  Buy/sell Cost  Shares Cost 
      $25   $100         4        $100  4      $25 
      $10   $100         10       $200  14     $14.3 
      $ 5   $100         20       $300  34     $8.8

Now let's say the price jumps to $15, your stocks are now worth $510, and he wants to sell some. I thank he is using $8.8/$15 to get .58 . At this point I am not sure how much he is selling of the stock value, it could be .58 or (1-.58 =.42), lets use .42. $510*.42= $214. $214/$15= 14 shares(rounded). He says he recalculates a new average cost.

Stock Monthly      Shares   Total Total  Average  
      Price installment  Buy/sell Cost  Shares Cost 
      $25   $100         4        $100  4      $25 
      $10   $100         10       $200  14     $14.3 
      $ 5   $30          6        $230  20     $11.5

The new average cost is $11.5.
Not sure if this is what he is doing or not.)

You said Value can never fall greater than 100% below cost but value can be any percentage above cost. So, there is a mathematical way of muting this disparity or of evening the percentages out. You probably learned this formula in High School algebra. Anyway, it is this that I use for the factor.
Could you post the mathematical formula, it has been a while since I hit the school books, thinks.

Concerning Synchrovest it does not use average buying price. My understanding of average buying price is this, you add up the prices that you bought stock at and divided by the number of prices. Synchrovest does not do this. It does use average cost. My understanding of average cost is, add up total cost and divide by total number of stocks. This is what Synchrovest does.
Note: in my spread sheet commissions are not included in total cost. But I feel they should be.

[2mc]

I will explain the gist of it. The crucial calculation is the true average cost which is based on LIFO. Since this is the case, a simple spreadsheet won't do. I had to program it because it required, when selling, looking back at purchases and picking off the number of shares at the different prices last to first. Then average cost will be based on the remaining shares and their combined actual cost.

On with the gist: I don't like AIM for 2 reasons: 1) Because it isn't truly based on average cost. Portfolio Control and SAFE are devices to 'guesstimate' the average cost. And, 2) because Portfolio Control can never go up when the true average cost can indeed go up. I also don't like Synchrovest for 2 reasons: 1) it doesn't use true average cost, but rather an average buying price and 2) it always buys into a stock no matter what the price. (When I say "I don't like" I don't mean it to sound as if there is something intrinsically wrong with either method nor am I trying to sound superior, it is just simply a personal preference. I believe AIM and Synchrovest have merit - that is, they work!)

I treat Cash and Stock Value equally. When Value is above cost I then multiply the Stock Value by a *factor* (I'll explain in a moment) and sell that amount. Conversely, when Value is below cost I then multiply Cash by a factor and buy that amount.

The factor is simply the ratio between value and cost - very similar to Synchrovest. However, I'm sure you will appreciate that there is a problem with ratios. For example, take two prices: 4 and 9. If one looks at the ratio in this way "4/9" then the answer is less than 50%. But, if one looks at the ratio in this way "9/4" then the answer is greater than 200%. So, you see there is a disparity in a simple ratio. Value can never fall greater than 100% below cost but value can be any percentage above cost. So, there is a mathematical way of muting this disparity or of evening the percentages out. You probably learned this formula in High School algebra. Anyway, it is this that I use for the factor.

So, every month there is a difference between value and cost and from month to month there will be a difference in the size of the ratios. So, on months where value and cost are close, not much money will be exchanged and on months where value and cost have diverged significantly, there will be a significant exchange of money. (This just seems to be in accord with commonsense.) And, since I'm in an IRA and in a program with virtually no commission (FolioFN) then I'm not concerned about tax consequences or commissions on small sales/purchases.

There's no tweaking involved. I believe the search for parameters is a never-ending fool's game. One will never find the optimum parameters for they will change. Even if you statistically find out what worked in recent history, the moment you try it you will be working with less than optimal parameters for they will have already changed. You can get a migraine trying to find the parameters. And, the search will *never* end.

So, I set up the account 50% cash and 50% stock. I guess you could tweak that, but why would anyone do that - unless they knew the future? And, if that were the case then they wouldn't be using AIM I can tell you that. <lol> I guess you could tweak what percentage of cash or stock the factor is multiplied by.

Balancing the portfolio within the Stock side of the program is very similar - if not the same - as Jibes has described.

The reason I use sector ETFs is that I don't know the future. I don't want to constantly try to figure out which stock is going up or which one of my holdings is in danger of going bankrupt. I also don't want to get into the game of following what's hot and what's not. So, I tried to get a diverse group of sectors that hopefully, by the time I retire, will have gone through at least one cycle of being hot and not. Hopefully, many sectors will go through several iterations. But, I don't know. I don't want to spend a whole lot of time on my retirement holdings - I'd rather spend that time trying to earn more to put into it.

So, there you have it in general terms. With a little investigation on your part you can fill in the gaps. (It's not rocket science, you know! <g>)

Matt

[karw]

Hi 2mc,

I have the impression that the ReBal machinery in your system is the same or very similar to the ReBal machinery that Jibes described.

However your adaptation of AIM based on average cost(Jibes that was what I use, I called it average price) is not completely clear to me.

Are you willing to describe it in more detail on this forum?

Thanks in advance, K

[jibes]

jibe.

I didn't set mine up with anthing in mind but I did select them at random. I would say try to use stocks from diferent sectors. This way the Re-Bal part would be much more effective.


Jibes
AIM Re Bal at:
http://jibes0.tripod.com/trendseeker.html

[jibes]

2mc

I see the mistake I made. I was fixing another mistake (same $3) in the table and mised the one in the text. Thanks.

Yes, your "volatility capture" idea sounds like a good one. It would be interesting to see how it would do as a AIM Re-Bal too.


Jibes
AIM Re Bal at:
http://jibes0.tripod.com/trendseeker.html

[jibe]

variety of stocks in portfolio

Is it better to have several "like" stocks in your portfolio or should they be more diverse?

Thanks

[2mc]

Sure.

S&P Sector 'Program'

XLB XLV XLP XLY XLE
XLF XLI XLK XLU


Misc. Sector 'Program'

DIA QQQ EFA IYR IYE
IYH IJK IJJ IWO IWW

Since FolioFN does not have every single ETF available (though a very high %), I had to make 2 substitutions for preferred ETFs that they did not carry.

Matt

[irwin]

Hi Matt could you please list the symbols for the whole group that you are using.
TIA
Irwin

[2mc]

Jibes,

On your 'school' page the following is written:

Next we go back to the TOTAL amount in line 5 and plug it into the Value column of AIM at line 10. In this case the number is $7756. Then as we work through AIM the amount for this month is an order to sell $293 worth of stock. We divide that by 4 and get -73 (Because we are selling, this is a negative number if we were buying it would be positive). Then we plug that number into line 7. at "FROM AIM =" and get the one just below it, line 8, "IDEAL VALUE =" of $1863 by subtracting 73 from $1939.
Each stock value is then compared to $1939 in the VALUE column. For example: On line 1 of XMSR which is $1720. Compare that to the "IDEAL VALUE =" and as you can see XMSR is down $146 from the ideal value of $1863 in line 8.

"$1863" should be $1866 in both cases. And, the sentence "Each stock value is then compared to $1939..." should read "Each stock value is then compare to $1866..."

It's interesting how independent minds can think similarly at similar times. Before your first posts I set up an account at FolioFN with a modified AIM (with a little Synchrovest twist thrown in) having a basket of ETFs that I rebalance. And, since I can buy/sell fractional shares of any size with no commissions (and because it's an IRA acct) I rebalance weekly and "AIM" monthly. I have set up two programs: one based on the S&P sector spiders - I have all nine - and one with 10 ETFs - Dow, cubes, foreign, real estate, natural resources, mid growth & value, small growth & value, and healthcare. Cash/Stock is 50/50 in both programs.

I call it my "volatility capture" method. I don't have to worry which sector is up and which is down. My only worry is that the whole market will be gone by the time I retire. And, if that happens, then I'll have other more pressing things to worry about.

I don't like strict AIM because it isn't truly based on average cost. Portfolio Control and SAFE are just devices - guestimates - to approximate average cost. It keeps one from having to keep track of which shares one is selling (and LIFO is the best way, by the way). So, in my modified AIM, money is flowing in or out every month based on an average cost 'factor.' So, I have nothing to 'tweak' - no historical testing to find parameters (that's a fool's game IMVHO). So, for me, I have an 'as automatic as I can make it' system. I don't have to evaluate stocks, worry whether or not they might go out of business, or worry whether they will continue their rhythm of volatility (they won't), or worry about finding the 'right' parameters. (and my 'no tweak' system beat AIM in Lichello's 10-8-5-4-5-8 sequence by 3.5 million dollars.)

I'd like to say 'great minds think alike' but I'm not sure I even have a 'less than average' mind. lol

Matt

[jibes]

Jibe

I haven't run any test scenarios yet, but what have you seen in terms of performance superiority vs. using just AIM?

Port 1: XMSR BA AEIS and RDC
AIM Re-Bal was +55% AIM on each stock done seperately garnered +23% as a group. Not too bad.

Port 2 ARTI PALM BWS and ZQK
AIM Re-Bal was +40% and AIM done on each stock totaled +24%.

I don't know how these Ports would have done if grouped together and AIMed without rebalancing so it's not a true comparison.

Those are the only two ports Iv'e done so far.

I notice in your example for february(re-bal), if you subtract the "value" column from the ideal value, the diff column seems to be off by $3(1863 - 1720 = 143). Am I missing a value somewhere?

I was waiting for someone to spot that. It's a mistake I made when doing that page. I need to fix it but I forgot to do that when updating my site.

Jibes

[jibe]

Question

I haven't run any test scenarios yet, but what have you seen in terms of performance superiority vs. using just AIM?

I notice in your example for february(re-bal), if you subtract the "value" column from the ideal value, the diff column seems to be off by $3(1863 - 1720 = 143). Am I missing a value somewhere?

Thanks

Robert

[jibes]

banjanxed

I don't have a spreadsheet that I want to distribute at this time. The one I have is just an aid to do the grunt work and would take a lot of explaining for folks to use it. One thing about this system is that it is very easy to make a mistake.
If you do it by hand take your time. If you have done AIM the regular way and used pen & paper you know what I'm talking about. It's easy to make mistakes.

Jibes

[banjanxed]

Hey Jibes,

I saw your post, and web site. Very interesting. Do you have a spreadsheet available to experiment with, or is this a totally manual (pen & paper) exercise? Cool concept, though. Good luck and keep on aiming!

banjanxed

[jibes]

Jibe

I got my nickname Jibes as a youth. Now I use it mostly for my handle on web sites and such.
Glad your interested in my AIM hybred. I was really surprized when I saw how powerfull it was.

I will be posting more examples on my web page and will be test more combinations (portfolios)

Jibes

[jibe]

Hey Jibes

I just started using AIM and am very interesting in trying your hybrid. I'll be on this board a lot.

It's always good to think outside the box.

Robert
(my "Investors Hub" user name is Jibe by the way) go figure.

[jibes]

Hi

I am just starting this AIM type site. I am quite excited by the profit potential of this AIM RE-Bal method.

I have added a twist to AIM and from what I have seen it looks quite exciting.

I know there will be plenty of questions so that is why I started this board.

jibes