ALSO DOWNLOAD THIS (right-click & save-as)--> | 3 year live performance of our timing models |

These are actuarial performance tables of all our JSE all-share index trading systems. To trade with these systems you purchase SATRIX40 or any TOP-40 Exchange Traded Fund (ETF). You can also add tabasco sauce to your trading by purchasing SATRIX-RESI, TOP40 CFD's, TOP40 warrants or SAFEX Index Futures to amplify gearing and/or performance.

These tables represent 15 years of performance from Jan 1997 to April 2012, of which the last 4-5 years have been LIVE with real customers. To just look at more recent 3-year live performance of some of the systems, download the "Live performance report" in the link at the top of this page.

MEASUREMENT & PERFORMANCE METRICS

When measuring the performance of the trading strategies, we look at the following metrics:

Trades = The amount of trades concluded during the period

Wins = How many trades resulted in a profit

Losses = How many trades resulted in a loss

Win% = What percentage of all trades were winners

Avg Trade = Average gain of all the trades (winners & losers) together

Avg Win = Average size of winning trades

Avg Loss = Average size of losing trades

Max Loss = maximum loss encountered

Avg drawdn = average peak-to-trough drawdown per trade

Max drawdn = maximum draw-down achieved across all trades

Pts up = percentage points accumulated in winning trades

Pts dn = percentage points lost in the losing trades

Net points = Pts up less Pts down

Gain/Loss = Pts up divided by points down

14yr TRI = total return of R1 re-invested in each trade over the entire period

CAGR = Compound annual growth rate achieved over the entire period

%Vested = how long the model kept you invested in the JSE

Sterling Ratio = compound growth divided by average draw-down (risk/reward measurement)

Tri Power = out performance of the JSE for the same measure of risk (1 = same as JSE)

The most important items we look at are the Win%, the Avg drawdn, the Gain/Loss ratio, the Sterling Ratio and the Tri Power. A high Win% is good for psychological reasons. Most non-professionals cannot stomach losses, especially strings of losses. The average draw-down is also important as large draw-downs cause discomfort during the course of the trade and are likely to trigger you into cutting the trade short in panic. Draw-downs can be viewed as the volatility or riskiness of the strategy. The gain/loss ratio shows how much more the system accumulates winning points than losing points - any value over 4 is deemed acceptable as then you claw back 4 points on your winning trades for each point you shed on the losing trades. The Sterling ratio is used by Hedge Funds to calculate risk adjusted return. High returns with gut-wrenching draw-downs will yield lower numbers than moderate returns with low draw-downs. Higher Sterling ratios are better than lower ones.

Finally the Tri-power is the total return achieved by the system divided by how long the system has been in the market (%Vested), then multiplied by 100 to compare returns similar to a fully vested buy+hold JSE strategy. We then divide this by the total returns (TRI) of the buy+hold over the same period. A number of 1 means the strategy achieved identical return for the same risk as a buy+hold. A value of 3 means the strategy achieved 3 times the buy+hold performance for the same risk. The higher this value the more the timing systems out-performs the buy+hold.

SOME NOTES BEFORE WE START

Why 15 years? Because that's the farthest back we could reconstruct market breadth data reliably for the JSE and many of our models use market breadth for their signalling. Most of these models were created from 2008 onwards meaning they included 11 years of back-testing. The last 3 years has therefore been "out of sample" or "live" performance - a good validation of the models' statistical robustness.

For comparison purposes, the JSE buy+hold strategy had a Total return (TRI) of 5.45 and a CAGR of 12.75% over the 15 year evaluation period. This means R1 invested in the system with profits re-invested in each subsequent trade, would now be R5.45 (444% growth over the 15 years or 12.75% compound growth.) The maximum draw-down was 45% giving a Sterling Ratio of 0.28. Quite clearly, the buy and hold strategy is not that risk-free as the industry would have you believe.

SECTION-A : INVESTMENT TIMING MODELS

Our main models in this category are the Composite SuperMODEL,the Long-Bond Yield curve (LBYC) timing model and the Cash/Equity ratio timing model. These are investment-grade models that look at economic fundamentals and financial ratios to determine favourable periods to be invested in the stock market, and unfavourable periods to avoid. They are very slow systems, trading 2-6 times per business cycle, where business cycles can range from 2-5 years in duration. These models typically achieve 4-7 times the performance of a buy+hold strategy for similar risk. A typical feature of these models is very high %Win rates. They almost all avoid recessions and bear markets.

The 1st SuperMODEL variant sells when the signal goes below -1 and the 2nd variant sells when the signal line goes below 0.

NOTE1 : For comparison purposes, the JSE buy+hold strategy had a TRI of 5.45 and a CAGR of 12.75% over the same period depicted above. The maximum draw-down was 45% giving a Sterling Ratio of 0.28. Quite clearly, the buy and hold strategy is not that risk-free as the industry would have you believe.

NOTE2 : Although this excersise goes back 15 years so that we can compare the investment models to all our other trading systems, bear in mind that SuperModel and LBYC history goes back some 35 years on the JSE so the above table is just a representation of more recent performance.

SECTION-B : LOW FREQUENCY, LONG TO MEDIUM TERM TRADING MODELS

These are low frequency, highly vested (long-term) and low vested (medium term) mechanical trading strategies that trade 1 to 2 times per year on average. The two highly vested systems on the left are in the market for 70% or more of the time and the 3 systems on the right of the table are in the markets less than 50% of the time.

The Zweig LazyBoy is promoted to our "newbie" clients or inexperienced traders since it is the most "kind" and "psychologically easier" system to trade due to its high win rate, low average loss, acceptable average draw-down, high gain/loss ratio and highest Sterling Ratio and Tri Power in the above group. The Turtle-S3 comes a very close second. You can read about the BITS twins over here.

SECTION-C : MEDIUM FREQUENCY, MEDIUM TO SHORT TERM TRADING MODELS

These are medium-frequency models that trade 2 to 4 times per year on average. The HedgeTrader is our most popular and powerful trader in this category delivering a class-leading Tri Power of 8.59 times the performance of the buy+hold with a very respectable 4.06 Sterling ratio. High win rates of 86% and stunning gain/loss ratios of 26:1 make this a popular choice among investment clubs. This system has the highest Total return (TRI) of all our systems (when re-investing profits from each trade into the next trade.) Although a maximum draw-down of 19.45% is hair-raising, it never booked a trade loss more than 5.76%. It is categorized as "highly vested" as it is in the market for 71% of the time as opposed to the others that are less than 50% of the time in the market.

The JSE SwissClock is our first ever timing model built for the medium term and lost popularity since the launch of HedgeTrader, but as you can see it is still a goodie. The Zweig Active is the more active variant of the Zweig Lazyboy which is discussed in the same research note. The STORM1 trades are hard-wired to 30 trading days (but HedgeTrader uses them for 23 days only) and are the shortest term trades here. It is amazing that a system that is in the market for a mere 37% of the time can deliver over 6 times the performance of the buy+hold for the same risk level.

SECTION-D : MEDIUM FREQUENCY, MEDIUM TO SHORT TERM TRADING MODELS

These high frequency systems trade on average 4 to 7 times per year. Their high frequency means they all have about 57% win rates which means only experienced traders that can manage losses, sometimes strings of losses, and who are disciplined from a money management perspective should participate. The old SwingTrader is shown, which stays out the market when HedgeTrader is out the market. This has been superseded by ULTRA. The McClellan Trader has yet to be documented on the web site, but it has a description in your JBAR report with its trading chart.

The Sincerity Trader, that uses simple re-introduction of back-to-back advances after an absence as its entry signal and a simple short-period Donchian lower channel as trailing stop, provides remarkable performance, outclassing the others in all the areas coloured green. It has the highest Stirling ratio of ALL our market timing systems. One interesting aspect of the Sincerity and McClellan traders is that they are self-preserving in bear markets, unlike SwingTrader that requires to consult HedgeTrader to determine if it is safe to trade or not.

NOTE : For comparison purposes, the JSE buy+hold strategy had a TRI of 5.45 and a CAGR of 12.75% over the same period depicted above. The maximum draw-down was 45% giving a Sterling Ratio of 0.28. Quite clearly, the buy and hold strategy is not that risk-free as the industry would have you believe.

These tables represent 15 years of performance from Jan 1997 to April 2012, of which the last 4-5 years have been LIVE with real customers. To just look at more recent 3-year live performance of some of the systems, download the "Live performance report" in the link below.

ALSO DOWNLOAD THIS (right-click & save-as)--> | 3 year live performance of our timing models |

These are actuarial performance tables of all our JSE all-share index trading systems. To trade with these systems you purchase SATRIX40 or any TOP-40 Exchange Traded Fund (ETF). You can also add tabasco sauce to your trading by purchasing SATRIX-RESI, TOP40 CFD's, TOP40 warrants or SAFEX Index Futures to amplify gearing and/or performance.

These tables represent 15 years of performance from Jan 1997 to April 2012, of which the last 4-5 years have been LIVE with real customers. To just look at more recent 3-year live performance of some of the systems, download the "Live performance report" in the link at the top of this page.

MEASUREMENT & PERFORMANCE METRICS

When measuring the performance of the trading strategies, we look at the following metrics:

Trades = The amount of trades concluded during the period

Wins = How many trades resulted in a profit

Losses = How many trades resulted in a loss

Win% = What percentage of all trades were winners

Avg Trade = Average gain of all the trades (winners & losers) together

Avg Win = Average size of winning trades

Avg Loss = Average size of losing trades

Max Loss = maximum loss encountered

Avg drawdn = average peak-to-trough drawdown per trade

Max drawdn = maximum draw-down achieved across all trades

Pts up = percentage points accumulated in winning trades

Pts dn = percentage points lost in the losing trades

Net points = Pts up less Pts down

Gain/Loss = Pts up divided by points down

14yr TRI = total return of R1 re-invested in each trade over the entire period

CAGR = Compound annual growth rate achieved over the entire period

%Vested = how long the model kept you invested in the JSE

Sterling Ratio = compound growth divided by average draw-down (risk/reward measurement)

Tri Power = out performance of the JSE for the same measure of risk (1 = same as JSE)

The most important items we look at are the Win%, the Avg drawdn, the Gain/Loss ratio, the Sterling Ratio and the Tri Power. A high Win% is good for psychological reasons. Most non-professionals cannot stomach losses, especially strings of losses. The average draw-down is also important as large draw-downs cause discomfort during the course of the trade and are likely to trigger you into cutting the trade short in panic. Draw-downs can be viewed as the volatility or riskiness of the strategy. The gain/loss ratio shows how much more the system accumulates winning points than losing points - any value over 4 is deemed acceptable as then you claw back 4 points on your winning trades for each point you shed on the losing trades. The Sterling ratio is used by Hedge Funds to calculate risk adjusted return. High returns with gut-wrenching draw-downs will yield lower numbers than moderate returns with low draw-downs. Higher Sterling ratios are better than lower ones.

Finally the Tri-power is the total return achieved by the system divided by how long the system has been in the market (%Vested), then multiplied by 100 to compare returns similar to a fully vested buy+hold JSE strategy. We then divide this by the total returns (TRI) of the buy+hold over the same period. A number of 1 means the strategy achieved identical return for the same risk as a buy+hold. A value of 3 means the strategy achieved 3 times the buy+hold performance for the same risk. The higher this value the more the timing systems out-performs the buy+hold.

SOME NOTES BEFORE WE START

Why 15 years? Because that's the farthest back we could reconstruct market breadth data reliably for the JSE and many of our models use market breadth for their signalling. Most of these models were created from 2008 onwards meaning they included 11 years of back-testing. The last 3 years has therefore been "out of sample" or "live" performance - a good validation of the models' statistical robustness.

For comparison purposes, the JSE buy+hold strategy had a Total return (TRI) of 5.45 and a CAGR of 12.75% over the 15 year evaluation period. This means R1 invested in the system with profits re-invested in each subsequent trade, would now be R5.45 (444% growth over the 15 years or 12.75% compound growth.) The maximum draw-down was 45% giving a Sterling Ratio of 0.28. Quite clearly, the buy and hold strategy is not that risk-free as the industry would have you believe.

SECTION-A : INVESTMENT TIMING MODELS

Our main models in this category are the Composite SuperMODEL,the Long-Bond Yield curve (LBYC) timing model and the Cash/Equity ratio timing model. These are investment-grade models that look at economic fundamentals and financial ratios to determine favourable periods to be invested in the stock market, and unfavourable periods to avoid. They are very slow systems, trading 2-6 times per business cycle, where business cycles can range from 2-5 years in duration. These models typically achieve 4-7 times the performance of a buy+hold strategy for similar risk. A typical feature of these models is very high %Win rates. They almost all avoid recessions and bear markets.

The 1st SuperMODEL variant sells when the signal goes below -1 and the 2nd variant sells when the signal line goes below 0.

NOTE1 : For comparison purposes, the JSE buy+hold strategy had a TRI of 5.45 and a CAGR of 12.75% over the same period depicted above. The maximum draw-down was 45% giving a Sterling Ratio of 0.28. Quite clearly, the buy and hold strategy is not that risk-free as the industry would have you believe.

NOTE2 : Although this excersise goes back 15 years so that we can compare the investment models to all our other trading systems, bear in mind that SuperModel and LBYC history goes back some 35 years on the JSE so the above table is just a representation of more recent performance.

SECTION-B : LOW FREQUENCY, LONG TO MEDIUM TERM TRADING MODELS

These are low frequency, highly vested (long-term) and low vested (medium term) mechanical trading strategies that trade 1 to 2 times per year on average. The two highly vested systems on the left are in the market for 70% or more of the time and the 3 systems on the right of the table are in the markets less than 50% of the time.

The Zweig LazyBoy is promoted to our "newbie" clients or inexperienced traders since it is the most "kind" and "psychologically easier" system to trade due to its high win rate, low average loss, acceptable average draw-down, high gain/loss ratio and highest Sterling Ratio and Tri Power in the above group. The Turtle-S3 comes a very close second. You can read about the BITS twins over here.

SECTION-C : MEDIUM FREQUENCY, MEDIUM TO SHORT TERM TRADING MODELS

These are medium-frequency models that trade 2 to 4 times per year on average. The HedgeTrader is our most popular and powerful trader in this category delivering a class-leading Tri Power of 8.59 times the performance of the buy+hold with a very respectable 4.06 Sterling ratio. High win rates of 86% and stunning gain/loss ratios of 26:1 make this a popular choice among investment clubs. This system has the highest Total return (TRI) of all our systems (when re-investing profits from each trade into the next trade.) Although a maximum draw-down of 19.45% is hair-raising, it never booked a trade loss more than 5.76%. It is categorized as "highly vested" as it is in the market for 71% of the time as opposed to the others that are less than 50% of the time in the market.

The JSE SwissClock is our first ever timing model built for the medium term and lost popularity since the launch of HedgeTrader, but as you can see it is still a goodie. The Zweig Active is the more active variant of the Zweig Lazyboy which is discussed in the same research note. The STORM1 trades are hard-wired to 30 trading days (but HedgeTrader uses them for 23 days only) and are the shortest term trades here. It is amazing that a system that is in the market for a mere 37% of the time can deliver over 6 times the performance of the buy+hold for the same risk level.

SECTION-D : MEDIUM FREQUENCY, MEDIUM TO SHORT TERM TRADING MODELS

These high frequency systems trade on average 4 to 7 times per year. Their high frequency means they all have about 57% win rates which means only experienced traders that can manage losses, sometimes strings of losses, and who are disciplined from a money management perspective should participate. The old SwingTrader is shown, which stays out the market when HedgeTrader is out the market. This has been superseded by ULTRA. The McClellan Trader has yet to be documented on the web site, but it has a description in your JBAR report with its trading chart.

The Sincerity Trader, that uses simple re-introduction of back-to-back advances after an absence as its entry signal and a simple short-period Donchian lower channel as trailing stop, provides remarkable performance, outclassing the others in all the areas coloured green. It has the highest Stirling ratio of ALL our market timing systems. One interesting aspect of the Sincerity and McClellan traders is that they are self-preserving in bear markets, unlike SwingTrader that requires to consult HedgeTrader to determine if it is safe to trade or not.

NOTE : For comparison purposes, the JSE buy+hold strategy had a TRI of 5.45 and a CAGR of 12.75% over the same period depicted above. The maximum draw-down was 45% giving a Sterling Ratio of 0.28. Quite clearly, the buy and hold strategy is not that risk-free as the industry would have you believe.

These tables represent 15 years of performance from Jan 1997 to April 2012, of which the last 4-5 years have been LIVE with real customers. To just look at more recent 3-year live performance of some of the systems, download the "Live performance report" in the link below.

ALSO DOWNLOAD THIS (right-click & save-as)--> | 3 year live performance of our timing models |