Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Strategies & Market Trends : A.I.M Users Group Bulletin Board -- Ignore unavailable to you. Want to Upgrade?

To: aptus who wrote (15174)	3/8/2001 2:28:23 AM
From: axp	Read Replies (1) \| Respond to of 18929

RE: After testing a pseudo-random selection of stocks, this algorithm produced worse results (compared to simple AIM BTB) in over 80% of the stocks tested. I also found that it performs slightly worse in most cases, which really says a lot for AIM BTB. However, the point of this discussion is to preserve cash in an extended decline, which of course would never happen to stocks we own :-). In that case, I found it does stretch the cash and increases the chance of having reserves to act on AIM buys when the bottom is put in. Missing these buys when a stock puts in a bottom really hurts returns so taking steps to avoid this are good, whatever they are. I'd also like to point out that if one's system is able to come up with cash when a stock's cash reserves are gone (via margin, pooled cash reserves, or borrowing from another stocks cash reserve) then you don't need a cash-extending mechanism because you'll have cash to buy on the AIM recommendations around the bottom. I probably shouldn't have brought up the WAIT%. I haven't done enough homework to back up the idea with statistics. I'm using it because for the stocks I'm betting a significant part of my portfolio on this modification made dramatic improvements in back-testing. For example, try it with CKFR from mid-1998 till now. Relative to Buy & Hold, I get +71% with BTB and +110% with a WAIT% in place. This is for a daily update frequency at the close. With weekly or monthly updates it doesn't help as much because most of the waiting is built into the less frequent updating.

To: aptus who wrote (15174)	3/8/2001 7:34:28 AM
From: OldAIMGuy	Read Replies (1) \| Respond to of 18929

Hi Mark, I would imagine that the time frame, the period of updates and whether a bear market had been included would greatly vary the results of such a test. I really don't worry too much about being out of cash. Think of the poor non-AIMer who is 100% invested all the time! No matter how the AIM users manages his cash, he still buys into any decline as deeply as AIM or his cash reserves allow. At that point we all know that any return to better prices is going to do AIM and ourselves a lot of good. We will have been more proactive (and probably more profitable) than almost any other type of investor. For me, there's the minor inconvenience of evaporating interest on the dwindling Cash Reserves. Since I take that interest as current income, I'm a much happier AIMer when the reserves are bloated! :-) Right now the family's on their "Tight Belt" allowances. (that's hard to do when facing up-coming Spring Breaks!) Best regards, Tom

To: aptus who wrote (15174)	3/8/2001 5:40:55 PM
From: labestul	Read Replies (1) \| Respond to of 18929

Hello Mark, The Money Spinner algorithm always uses the previous transaction price as a starting base. It does not matter if the transaction was a buy or a sell. Also the new buy price (for example) is compared not to a percentage of the previous buy price but to a fixed absolute deviation from the previous transaction price. Are you coming to AIM 2001? If so, maybe we can get together and discuss this in more detail. I am of course assuming that you have no objections to a concurrent libation. I also would like to discuss your "volumizer" at that time. I will write a subsequent post comparing the results of strict Money Spinner by the book to strict AIM by the book on some select stocks. The first I plan to show is UOPIX. If there is any other particular stock you would like to see please let me know. Perhaps this is a good point at which to spell out (simply I hope) the Money Spinner algorithm. Before doing so let me point out that I am not necessarily advocating it over AIM by the book. In fact in practice I still only use strict AIM by the book but my philosophy has been influenced by the Money Spinner. One starts a Money Spinner account exactly like an AIM account. That is 50% of the money purchases stock and 50% goes to cash. Of course more modern AIM will use different cash levels but the Money Spinner came out in 1980 just three years after the original AIM book. One further wrinkle is that one should buy a stock whose price is such that about 1,000 shares (or a multiple thereof) are purchased. In the rest of this description I will assume that we purchased 1,000 shares and will speak of buying or selling 100 shares at a time. If the purchase had been 3,000 shares for example then substitute 300 shares for 100 shares in the following description. Basically the Money Spinner is AIM with a slight twist. First, using 10% buy and sell safes, calculate what the price per share would be if AIM directed you to sell 100 shares. Similarly calculate the price for buying 100 shares. Then put in GTC orders and wait until either a buy or a sell takes place. Note that there is no recognition of minimum transaction size other than that implicit in the 100 share transaction minimum which is 10% of the initial market value but not necessarily of subsequent transactions. This is one reason why I don't recommend Money Spinner by the book. There is one more step. You must calculate minimum buy and sell intervals. The minimum sell interval (MSI) is equal to the next sell price in your GTC sell order minus the initial purchase price of the stock. Similarly the minimum buy interval (MBI) is equal to the initial purchase price minus the next buy price in your GTC buy order. The MSI and MBI are calculated when you set up your Spinner and do not change. Now you wait until one of the GTC orders is acted upon. You immediately cancel the other GTC order (i.e. the one not acted on) and then issue two new GTC orders, one a sell and one a buy. How do you set these new GTC orders? You calculate new buy and sell prices for 100 shares just as before. Again you use strict AIM to do this calculation. Now the next step is to compare the buy and sell price to the transaction price that was just used. The new sell price must be greater than this transaction price by at least the MSI. If it is then enter a GTC order to sell 100 shares at that price. If it is not then you set the new sell price to the previous transaction price plus the MSI. Then you determine the number of shares (which will be greater than 100) that AIM would direct you to sell if the current price were this new price. Now you enter a GTC sell order for this calculated number of shares at the previous transaction price plus the MSI. Similarly if the new buy price that you calculated above is less than the previous transaction price by at least the MBI then enter a GTC order to buy 100 shares at that price. If not then set the new buy price to the previous transaction price minus the MBI. Then determine how many shares AIM would direct you to buy (again this will be more than 100 shares) if that were the current price. Enter a GTC buy order for that number of shares at a price equal to the previous transaction price minus the MBI. It should be noted that just like in AIM, every time more shares are purchased half of the purchase cost is added to the portfolio control (which is called the flow limit in Money Spinner parlance). Take Care, Barry