The following table however shows both 1:1 gearing, the actual Index return (mid prices to mid prices) and 1.5 and 2 times gearing factors.

Note: I do not have the Dividend Yield data for the specified date ranges or the Interest rates applicable during the 1966-75 period, so assumptions have been made. Interest rates shown for 1985-2000 are those applicable as of January 1st of each year.

The estimated IG Buy and Sell prices include an assumption of a 0.15% spread charge against both buy and sell values which appears relatively realistic against current spread charge ranges.

It is assumed that interest earnt against cash deposits can be identified such that the net return against such a deposit would match the current base lending rate. In practice, with the current base rate at 6%, there are accounts around that are paying 7.15%, so for a basic rate taxpayer, compares reasonably well. Under such conditions the returns shown below against the tracker would be net of all dealing costs and tax.

Note that the data for Dec 75 (NYSE) is included. With such being present it shows that the combined periods of 1984-2000 and 1966-75 have a better than Index return against the 1.5 geared position. However if that December data is omitted and then the overall combined shows values of

6.228399
6.0416473
5.065866017

respectively. Principly arising out of the 1966-75 period where no growth in the Index was apparent (50:50 cash/tracker part didn't cover the FSBS interest premium costs). Interestingly though, the method still shows that following Dec 75, the method returned to beating the Index again. In contrast, during the 1984-2000 years, considerable outperformance is apparent.

The 1966-75 period was the worst period that I could find data for (bear period) and was included so as to give an indication that things might not always be as smooth as the last 25 years of a Bull run, but that even when sustained down periods are encountered, a recovery back to bettering the Index overall will probably occur in the longer term.

If the 1975-1984 period was included in this set (also a Bull period), then overall the 1.5 geared approach would be clearly indicated as a principly winning strategy. As an indication, the NYSE presently stands at around the 650 level which is 12.88 times its original value of Jan 1966 which is around 11.25% average p.a. (exc dividends) which betters bank rates and therefore would have resulted in our strategy outperforming the index overall (remember that to beat the Index, all we have to do is beat the Base Interest Rate).

For reference, over the 10 years Jan 1990 - Jan 2000, the index grew by an average of 11% p.a. Combined with 2% p.a. dividends and in view of a 8.33% average interest rate, I would have expected to see our method better the index by around 2.35% p.a. So with an average of 11% growth, 2% dividend and 2.35% outperformance = 15.35% p.a. NET return (after all costs). Which equates to a growth of 4.17 times the original investment over that period (yield included). Have a hunt around and see how many Unit Trusts or Investment Trusts offer similar returns (don't forget to add all dealing costs, charges and also allow for tax - as a guide for basic rate taxpayers, a UT or IT will need to return at least 5.0 times the original investment to compare with this method. Whilst some can be found, they are in the minority and probably have significantly high volatilities. If you do find one that does appear to consistently outperform this method, then why not replace the 50:50 cash:tracker part of the method with that Investment Trust and use it to serve the main Index Tracker and you'll probably get even higher returns than that Investment Trusts return).

Overall for the combined period, the best investment was the 1:1 tracker! Whilst cash did better the 1.5 Index tracker over the sample combined period the period between the two sample ranges (1975-85) would however have pushed shares back into the lead again.

If you can pre-empt the markets performance in any year, then a sliding Gearing factor would probably serve best e.g. ranging from Cash based to 2 times gearing according to anticipated market performance over the forthcoming year. Similarly picking the right Index to track would also help increase profits.

You may ask, but I've heard 2/3rds of profits over the long term arise from dividend re-investment, your chart doesn't include such re-investments, if it had, wouldn't that have enhanced the overall profits to above the 1:1 tracker? Well it works both ways, I estimate that re-investment of dividends would have produced figures of 7.15 times original for the 1985-2000 period and 1.31648 to the 1966-1975 period, combining producing 9.4128 times the original investment over the combined period for the mid-mid (zero cost) based example tracker. However, the 1:1 tracker also used a fixed £10 per point stake level throughout, if that had been adjusted for the amount of cash available at the start of each year, a correspondingly higher overall return would have been produced. As an example the last year of the 1999-2000 period, a cash reserve of £68689 was available against an index level of 5909 - so the per point stake level could have been over 16% higher (e.g. £11.62 per point instead of 10) and produced a return some 16.2% higher than the return shown in the 1:1 column for that year. Similarly such variable per point levels could have been reflected throughout all other years.