shown in Table 1.11. We showed earlier that the equivalent annual gain for the transactions increased dramatically as we shortened the length of time for the transaction down to 12 weeks, but that it increased only marginally as we moved to transactions which lasted only 2.6 weeks. The same effect appears to carry through when we compound the gains by reinvestment into the succeeding transaction, since the final column where these values are displayed shows a fall-off from 55,100% for the 12-week transactions to 235% for the 2.6-week transactions. However, these two figures are not directly comparable, since the last one applies to a period of only 3.1 years, and not 18 years as do all the previous figures.
To bring the 3.352 gain factor over 3.1 years to one over 18 years we first bring it to an annual gain of 1.4772 by using the method we discussed at the beginning of this chapter. By raising this value to the power 18, we get the equivalent gain over 18 years, which is 1122. In percentage terms this is 112,100%.
Note that this theoretical gain of 112,100% has been obtained with our money working for us only part of the time. Taking Grand Metropolitan shares as an example, we can work out how much of the time we were invested by multiplying the average length of the particular trend, i.e. 45 weeks, 12 weeks and 2.6 weeks respectively, by the number of such trends that occurred over the 936-week period. Taking the 12-week trends as an example, there were 41 of these in the 936-week period. This means we were invested for 41 x 12 = 492 weeks out of the possible 936. During the rest of the time we would have been earning at the rate of what now appears to be the positively miserly 6% per annum, or thereabouts, that could have been obtained in the money market over this period since 1978. Quite obviously, therefore, we have to try to reach the position where our money is invested in rising shares 100% of the time, or as close to that as possible. If we can do that then we will obviously improve the gain made over an extended series of transactions enormously.
FROM THE THEORETICAL TO THE ACTUAL
So far in this chapter we have made two basic assumptions that we have not yet qualified: firstly, that we achieve perfect timing of our buying and selling operations, and secondly that none of the investments go wrong and lose us money. We ignored this aspect simply to develop the argument that the gains which can be made from successive short-term investment outweigh by tens, hundreds or even perhaps thousands of times the gain which is made by a simple buy and hold strategy. These astronomical gains must therefore be considered to be theoretical in nature. It will be totally impossible, whatever method we use, to buy in at the bottom and sell at the top of these trends consistently, year in and year out. We might do it once or twice over the course of say a dozen transactions, but even that would be lucky. Furthermore, it is impossible to predict with 100% certainty that a trend will continue to rise so as to give us a guaranteed profit from every transaction. There will be occasions when random influences will cut short a trend, reversing its direction at such a speed that we cannot avoid a loss.
Table 1.12 Compounded gains made from 20 consecutive transactions for different gain levels per transaction. It is assumed that there are 16 winners and 4 losers. The loss for the losers is the same as the gain for the winners
Being realistic, therefore, we have to downgrade our expectations for profit from the levels we have been using for the calculations in the previous tables. Firstly, we will make the assumption that the maximum gain per transaction after dealing costs will be 20%, i.e. using a round number slightly higher than the 17.94% which we showed previously would follow from perfect timing. Secondly, we will make the assumption that we are correct eight times out of ten, and for simplicity, when we are wrong, we will lose the same percentage that we make when we are correct. Thirdly, we will assume that the investor makes ten consecutive transactions, reinvesting the proceeds each time. Table 1.12 shows the resulting gains for such a series of investments where the gain (and loss), after dealing costs are taken into account, varies from 20% down to 1% per transaction.
These compounded gains run from 657%, i.e. multiplying our starting capital by 7.57 when we achieve a 20% gain per winning transaction and a 20% loss for losing transactions, down to 12.6% where we only achieve a 1% gain or loss. To double our capital over these 20 transactions we need to reach the level of just over 6% per transaction. Note the improvement made for each additional 1% that can be squeezed out. Thus the investor reaching 7% per transaction will do 20% better overall than the investor reaching 6%.
Most of this book is dedicated towards improving the timing of buying and selling operations to such an extent that we should be able to capture gains of around 8 to 10% (after dealing costs) from each of those transactions which we have correctly forecast, while restricting our losses to similar levels from each of those transactions where the forecast goes wrong. From Table 1.12 it can be seen that this means that our capital will increase by a factor of two and a half to three times after 20 such transactions. The remainder of this book is dedicated towards improving the performance of investors so that they can make these extra few percentage points out of each rise. Investors will be able to concentrate on the shorter-term trends such as the 12-week trends we have been discussing for Grand Metropolitan. Techniques will be shown that enable the best shares to be selected to take advantage of these short-term trends. Techniques will also be shown that enable the investor to buy in very early in the life of the uptrend and sell not too far down from the end of the trend. It is suggested that investors develop a five-year horizon. In this time period, there will be between 20 and 25 transactions in just one share if trends of the order of 12 weeks are used. The investor who wishes to become more deeply involved can be invested simultaneously in a number of shares, subject to the restrictions mentioned in the final chapter.
In summary we can make the following points:
Share prices consist of upward and downward trends of varying lengths of time.
These trends fall into various categories, including those that last on average less than three weeks, those that last on average about 12 weeks and those that last on average just under one year.
Average investors should make gains of about 10% out of trends which last on average for 12 weeks.
By compounding such gains, average investors should be able to multiply their capital by about three over a series of 20 such transactions.
Channel analysis will improve performance so that gains of many tens of times are possible over a long term if a full investment strategy is pursued as far as is practicable.
Chapter 2. The Nature of Share Price Movement
INTRODUCTION
There has always been controversy over the way in which share prices move over the course of time, with chartists maintaining that prices can be predicted to a certain extent because historical patterns in the charts of share prices tend to recur from time to time. These methods of analysis rest heavily on the recognition of the start of a pattern formation so that the subsequent movement can be anticipated. On the other hand, the fundamentalists believe that the key to investment success lies in such factors as the way in which a company is managed, the quality and appeal of its products, and the strength or weakness of its balance sheet.
They believe for the most part that chartist techniques are just mumbo-jumbo and that the past history of share price plays no part in the future movement. If pressed about the nature of share price movement, many fundamentalists would state that they believe that share prices move on a random basis and therefore cannot be predicted. In doing this, they ignore the obvious corollary: if prices move randomly, there is no advantage in studying the fundamentals of any company since the random share price will bear no relationship to these fundamentals.
While
