Jack wrote:
Some people feel that it doesn't really matter where on the price cycle one starts AIMing, and for a stock that moves as follows...10,8,6,4,6,8,10...the most profitable entry point is 8. That could be true, but personally, I don't buy that. Since AIM is a buy low / sell high system, why would I knowingly start at 8? It seems illogical to me.
This raises a lot of questions with me. Firstly this looks a lot like hypothetical sequence which Lichello gives in his book however his sequence is 10, 8, 5, 4, 5, 8, 10 with a 5 instead of 6.
From where do you get the claim that 8 might be the most profitable entry point and even if it is which 8 do you mean? The one in the falling 10, 8, 6 pattern or the one in the rising 6, 8, 10 pattern. These two different 8's might yield different results.
Also what is meant by most profitable? This clearly begs the definition!!!
I will not go through the 10, 8, 6, 4, 6, 8, 10 sequence since I have already analyzed Lichello's 10, 8, 5, 4, 5, 8, 10 sequence however I believe the two should give similar results.
Let's assume that the price of 10 is in effect on January 1, 1992 and that next month on February 1, 1992 the new price is 8 etc. Following the tables from pages 64 through 71 inclusive we see that every January 1st and July 1st the price returns to 10.
Let us assume that we check once a month on the 1st and enter our transactions accordingly. Also let us assume standard 10% buy and sell safes but otherwise no minimum transaction sizes. Finally let us assume neither any commission fees nor any interest income on our cash.
Finally let us note that the last entry in Lichello's table corresponds to a date of July 1, 1999 using my above dating scheme.
If I now go through and simulate all trades assuming that I start with 10,000 on Jan. 1, 1992 and invest 5,000 (i.e. half) into the hypothetical stock. My simulation more or less duplicates Lichello's. The only differences between mine and his is that I include no interest income. Also the number of shares in some of my transactions differ from his by one share. This is due no doubt to rounding. In any event mine is close enough.
As of July 1, 1999 my total of cash and market value of stocks is $1,147,646 which implies a 5.41% overall rate of return per month on my investment over the whole period.It is however more interesting to compute the actual IRR on the moneys actually invested in the stock. This is what some here call ROCAR I believe. This value 7.30% per month.
If I repeat the same simulation (i.e. 5,000 cash and 5,000 in stocks) but starting on February 1, 1992 and looking at the results on July 1, 1999 I can calculate the same results as I had done for the January 1, 1999 start.
I have done this for each of the start dates of January 1, 1992 through June 1, 1992 inclusive. These correspond to per share prices of 10, 8, 5, 4, 5, and 8 respectively. Here are the results.
DATE / TOTAL VALUE / OVERALL RETURN RATE / ROCAR / PRICE
Jan 1 / $1,147,646 / 5.41% / 7.30% / 10
Feb 1 / $1,034,492 / 5.35% / 7.62% / 8
Mar 1 / $ 832,750 / 5.15& / 8.68% / 5
Apr 1 / $ 787,444 / 5.15% / 10.73% / 4
May 1 / $ 786,258 / 5.21% / 9.21% / 5
Jun 1 / $ 789,855 / 5.27% / 7.51% / 8
It's clear here that in terms of maximum overall return it was better to start at 10 and not 8 or any other value.
Or is it????
Note that by starting at 4 the ROCAR was clearly the highest by far. Why then is the overall return not so low compared with starting at 10 or 8 or even 5 (descending)?
The answer is that we had much less invested in stocks on average. Starting at 10 we bought more shares initially but starting at 4 we could only sell shares.
To make this a fair comparison the proportion of cash should have been much less starting at 4 than starting at 10. Thus comparing the total results is not fair and is rather like comparing apples to oranges.
Of course this is all very well in theory and no doubt we would all agree that there should have been less cash starting at 4 BUT ... how can we tell what proportion should be in cash??? This is another question entirely.
Thus I have shown at least two different definitions of 'better' as in which starting point is 'better'.
Clearly starting at 4 gives the best return on capital at risk whereas starting at 10 gives the best overall return in this artificial example. In practice I would always go with the method that gave me the best ROCAR and adjust overall returns by carefully watching my cash portion.
Barry Savage
mailto:bv@bsavage.net |
